Skip to main content

Package for curating data annotation efforts.

Project description

pyAnno 2.0 is a Python library for the analysis and diagnostic testing of annotation and curation efforts. pyAnno implements statistical models for inferring from categorical data annotated by multiple annotators

  • annotator accuracies and biases,

  • gold standard categories of items,

  • prevalence of categories in population, and

  • population distribution of annotator accuracies and biases.

The models include a generalization of Dawid and Skene’s (1979) multinomial model with Dirichlet priors on prevalence and estimator accuracy, and the two models introduces in Rzhetsky et al.’s (2009). The implementation allows Maximum Likelihood and Maximum A Posteriori estimation of parameters, and to draw samples from the full posterior distribution over annotator accuracy.


pyAnno is licensed under a modified BSD license (2-clause). For more information, see



The documentation is hosted at .


  • Pietro Berkes (Enthought, Ltd.)

  • Bob Carpenter (Columbia University, Statistics)

  • Andrey Rzhetsky (University of Chicago, Medicine)

  • James Evans (University of Chicago, Sociology)

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pyanno-2.0.2.tar.gz (292.5 kB view hashes)

Uploaded source

Built Distributions

pyanno-2.0.2.macosx-10.5-i386.exe (296.3 kB view hashes)

Uploaded any

pyanno-2.0.2-py2.7.egg (377.8 kB view hashes)

Uploaded 2 7

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page